Hepatitis C virus (HCV) is the major cause of chronic hepatitis with a significant risk of end-stage liver cirrhosis and hepatocellular carcinoma. (See Di Bisceglie, A. M., (1997) Hepatology 26(3 Suppl. 1): 345-385). HCV belongs to the family Flaviviridae, which consists of three genera: flaviviruses, pestiviruses, and hepaciviruses. In the absence of both a suitable small animal model and a reliable in vitro infectivity assay for HCV, potential antiviral drugs are initially tested using a related pestivirus, bovine viral diarrhea virus (BVDV). BVDV in vitro infectivity assays have been used to demonstrate that alkylated iminosugar derivatives containing either the glucose analogue 1,5-Dideoxy-1,5-imino-D-glucitol, also called deoxynojirimycin or “DNJ,” or the galactose analogue 1,5-Dideoxy-1,5-imino-D-galactitol, also called deoxygalactonojirimycin or “DGJ,” are potent antiviral inhibitors. (See Durantel, D., et al., (2001) J. Virol. 75(19): 8987-98).
DNJ derivatives are antiviral inhibitors at least partially because they inhibit ER α-glucosidases I and II. These enzymes remove three glucose residues that form part of N-glycan precursors, which are transferred en bloc to nascent glycoproteins in the ER. ER α-glucosidase inhibition prevents the formation and subsequent interaction of monoglucosylated, N-linked oligosaccharide-containing glycoproteins with the ER resident chaperones calnexin and calreticulin. (See Bergeron, J. J., et al., (1994) Trends Biochem. Sci. 19(3): p. 124-8; Peterson, J R., et al., (1995) Mol. Biol. Cell 6(9): 1173-84). Calnexin interaction is crucial for the proper folding of many host and virus encoded glycoproteins, including the envelope glycoproteins of BVDV and HCV. (See Branza-Nichita, N., et al., (2001) J Virol. 75(8): 3527-36; Choukhi, A., et al., (1998) J. Virol., 1998 72(5): 3851-8). Misfolding of BVDV envelope glycoproteins may lead to an impaired secretion of virions from infected cells.
Previous experiments have shown that the antiviral effect of the long alkylchain derivative N-nonyl-DNJ (NN-DNJ) is more pronounced than that of the short alkylchain derivative N-butyl-DNJ (NB-DNJ), although the latter achieves a more effective ER α-glucosidase inhibition in cellulo. (See Durantel, D., et al., (2001) J. Virol. 75(19): 8987-98). In addition, long alkylchain DGJ-derivatives which are not recognized by and do not inhibit ER α-glucosidases, also show potent antiviral activity. Therefore, ER α-glucosidase inhibition does not directly correlate with the observed antiviral effect and is ruled out as the sole antiviral mechanism.
The additional mechanism of action is apparently associated with the length of the alkyl sidechain, as the shortchain N-butyl-DGJ (NB-DGJ) shows no antiviral activity, whereas the long alkylchain derivative NN-DGJ is a potent inhibitor. It is, however, not associated with a detergent-like effect of the amphiphilic, alkylated iminosugar derivatives, as the structurally similar detergents n-octyl glucoside (n-OG) and n-nonyl glucoside are not antiviral in in vitro BVDV infectivity assays. (See Durantel, D., et al., (2001) J. Virol 75(19): 8987-98). Drug treatment affects the dimerization of viral membrane glycoproteins and alters the membrane glycoprotein composition of secreted BVDV virions, but does not influence either viral RNA replication or protein synthesis.
Because of their ER α-glucosidase inhibitory activity, both long and short alkylchain DNJ-derivatives are antiviral against flaviviruses like Dengue virus (DENV) and Japanese Encephalitis virus (JEV). (See Courageot, M. P., et al., (2000) J Virol. 74(1): 564-72; Wu, S. F., et al., (2002) J. Virol., 76(8): 3596-604). In contrast, DGJ-derivatives show no antiviral activity against flaviviruses, but the long alkylchain DGJ-derivatives are potent inhibitors of pestiviruses. Unlike the closer related pesti- and hepaciviruses, flaviviruses do not contain p7 or an equivalent small membrane spanning protein. As such, DNJ- and DGJ-derivatives may inhibit viral replication by inhibiting p7 or equivalent proteins.
The HCV positive stranded RNA genome encodes a single polyprotein precursor of approximately 3000 amino acid residues. The polyprotein is co- and post-translationally processed by viral and cellular proteases to produce the mature structural and non-structural proteins C, El, E2, p7, NS2, NS3, NS4A, NS4B, NS5A and NS5B (reviewed in Reed, K. E. and C. M. Rice, (2000) Curr. Top. Microbiol. Immunol. 242: p. 55-84.), with a potential further F protein arising from a ribosomal frameshift in the N-terminal region of the polyprotein. (See Xu, Z., et al., (2001) EMBO J. 20(14): p. 3840-8). Although most cleavages in the polyprotein precursor are efficiently completed during or immediately after translation, cleavages are delayed at the E2/p7 and p7/NS2 sites, leading to the production of an E2-p7-NS2 precursor. Furthermore, processing between E2 and p7 is incomplete resulting in the production of E2-p7 as well as E2 and p7. (See Lin, C., et al., (1994) J. Virol. 68(8): p. 5063-73; Mizushima, H., et al., (1994) J. Virol. 68(10): p. 6215-22).
Because a cell culture model for HCV replication is unavailable, information about HCV replication is derived by using a BVDV cell culture model. As such, most functional data about p7 are derived from studying BVDV p7, a 70 amino acid protein very similar to HCV p7. Functional data about BVDV p7 has been obtained by introducing mutations into an infectious cDNA clone of BVDV. An in-frame deletion of the entire p7 gene does not affect BVDV RNA replication, but does lead to the production of non-infectious virions. However, infectious viral particles can be generated by complementing p7 in trans. (See Harada, T., N. Tautz, and H. J. Thiel, (2000) J. Virol. 74(20): 9498-506), which suggests that the pestivirus p7 is essential for the production of infectious progeny virus.
The HCV p7 protein is a 63 amino acid peptide which has been shown to be a polytopic membrane protein that crosses the membrane twice and has its N- and C-termini oriented towards the extracellular environment. (See Carrere-Kremer, S., et al., (2002) J. Virol. 76(8): 3720-30). As such, the p7 protein has been shown to include two transmembrane domains. The N-terminal transmembrane domain includes amino acids from about position 10 to about position 32 and the C-terminal transmembrane domain includes amino acids form about position 36 to about position 58. Although the amino acids within the two transmembrane domains are somewhat variable among all HCV strains, for reported strains, a majority of amino acids within the transmembrane domains, (typically greater than about 70%), are members of a hydrophobic group characterized as F, I, W, Y, L, V, M, P, C, and A. The two transmembrane domains are linked by three non-hydrophobic amino acids, (K or R, G, R or K), and a consensus sequence for p7 is ALENLVVLNAASAAGHTGILWFLVFFCAAWYVKGLRVPGATYSLLGLWPLLL LLLALPQRAYA (SEQ ID NO.: 1). (See Carrere-Kremer, S., et al., (2002) J. Virol. 76(8): 3720-30).
Subgenomic replicon studies have shown that the p7 of HCV is not necessary for genome replication. (See Lohmann, V., et al., (1999) Science 285(5424): p. 110-3; Pietschmann, T., et al., (2001) J. Virol. 75(3): p. 1252-64), but its role in infectious virus production is unknown. HCV p7 has been suggested to be a putative member of a group of small proteins known as viroporins, which mediate cation permeability across membranes and are important for virion release or maturation. For some viroporins, the transmembrane domain has been shown to be sufficient to form membrane channels. (See Duff, K. C. and R. H. Ashley, (1992) Virology 190(1): p. 485-9; Fischer, W. B., et al., (2001) Eur. Biophys. J. 30(6): p. 416-20).